Scholarlink

Background <p>The comprehensive investigation into the aerodynamic behaviour of high-rise composite plan-shaped buildings uses a hybrid approach that combines Computational Fluid Dynamics (CFD) with machine learning techniques.</p> Purpose <p>The study focuses on analyzing complex wind flow interactions, pressure distributions, and structural responses under varying wind incidence angles ranging from 0° to 90° at 15° intervals.</p> Methods <p>Four building models with different upper-section geometries—square, circular, hexagonal, and octagonal—are designed and evaluated. Simulations are conducted using the ANSYS CFX solver with a 5% turbulence intensity and the k–𝜀 turbulence model to capture steady-state wind dynamics. The study also assesses structural symmetry and identifies regions prone to vortex shedding and aerodynamic instabilities.</p> Result <p>To enhance prediction efficiency, a supervised machine learning model is trained on simulation data to forecast pressure coefficients and streamline patterns, significantly reducing computational costs without compromising accuracy. The integration of CFD with machine learning enables a deeper understanding of aerodynamic performance across diverse geometrical configurations.</p> Conclusion <p>The findings provide practical guidance for architects and structural engineers in designing aerodynamically efficient and sustainable high-rise structures. This research advances smart urban planning, energy efficiency, and structural safety in high-rise building design, addressing key challenges in modern architecture and engineering.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Numerical Simulation and Machine Learning-Based Analysis of Aerodynamic Characteristics in High-Rise Composite Plan-Shaped Buildings

  • 期刊论文

  • 发表日期：

    2026年5月25日

查看全文
  • Dharmendra Kumar,
  • Pradeep K. Goyal,
  • Ritu Raj

摘要

Background

The comprehensive investigation into the aerodynamic behaviour of high-rise composite plan-shaped buildings uses a hybrid approach that combines Computational Fluid Dynamics (CFD) with machine learning techniques.

Purpose

The study focuses on analyzing complex wind flow interactions, pressure distributions, and structural responses under varying wind incidence angles ranging from 0° to 90° at 15° intervals.

Methods

Four building models with different upper-section geometries—square, circular, hexagonal, and octagonal—are designed and evaluated. Simulations are conducted using the ANSYS CFX solver with a 5% turbulence intensity and the k–𝜀 turbulence model to capture steady-state wind dynamics. The study also assesses structural symmetry and identifies regions prone to vortex shedding and aerodynamic instabilities.

Result

To enhance prediction efficiency, a supervised machine learning model is trained on simulation data to forecast pressure coefficients and streamline patterns, significantly reducing computational costs without compromising accuracy. The integration of CFD with machine learning enables a deeper understanding of aerodynamic performance across diverse geometrical configurations.

Conclusion

The findings provide practical guidance for architects and structural engineers in designing aerodynamically efficient and sustainable high-rise structures. This research advances smart urban planning, energy efficiency, and structural safety in high-rise building design, addressing key challenges in modern architecture and engineering.